-
Notifications
You must be signed in to change notification settings - Fork 30
/
main.py
184 lines (151 loc) · 6.5 KB
/
main.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
import torch
import torch.nn as nn
import torch.optim as optim
import sys, os
import visdom
import torchvision.models as models
from torch.utils.data.sampler import SequentialSampler
from torch.utils.data import DataLoader
from dataloader.imfol import ImageFolder
from utils import transforms as custom_transforms
from models import scribbler, discriminator, texturegan, define_G, weights_init, \
scribbler_dilate_128, FeatureExtractor, load_network
from train import train
import argparser
def get_transforms(args):
transforms_list = [
custom_transforms.RandomSizedCrop(args.image_size, args.resize_min, args.resize_max),
custom_transforms.RandomHorizontalFlip(),
custom_transforms.toTensor()
]
if args.color_space == 'lab':
transforms_list.insert(2, custom_transforms.toLAB())
elif args.color_space == 'rgb':
transforms_list.insert(2, custom_transforms.toRGB('RGB'))
transforms = custom_transforms.Compose(transforms_list)
return transforms
def get_models(args):
sigmoid_flag = 1
if args.gan == 'lsgan':
sigmoid_flag = 0
if args.model == 'scribbler':
netG = scribbler.Scribbler(5, 3, 32)
elif args.model == 'texturegan':
netG = texturegan.TextureGAN(5, 3, 32)
elif args.model == 'pix2pix':
netG = define_G(5, 3, 32)
elif args.model == 'scribbler_dilate_128':
netG = scribbler_dilate_128.ScribblerDilate128(5, 3, 32)
else:
print(args.model + ' not support. Using Scribbler model')
netG = scribbler.Scribbler(5, 3, 32)
if args.color_space == 'lab':
netD = discriminator.Discriminator(1, 32, sigmoid_flag)
netD_local = discriminator.LocalDiscriminator(2, 32, sigmoid_flag)
elif args.color_space == 'rgb':
netD = discriminator.Discriminator(3, 32, sigmoid_flag)
if args.load == -1:
netG.apply(weights_init)
else:
load_network(netG, 'G', args.load_epoch, args.load, args)
if args.load_D == -1:
netD.apply(weights_init)
else:
load_network(netD, 'D', args.load_epoch, args.load_D, args)
load_network(netD_local, 'D_local', args.load_epoch, args.load_D, args)
return netG, netD, netD_local
def get_criterions(args):
if args.gan == 'lsgan':
criterion_gan = nn.MSELoss()
elif args.gan == 'dcgan':
criterion_gan = nn.BCELoss()
else:
print("Undefined GAN type. Defaulting to LSGAN")
criterion_gan = nn.MSELoss()
# criterion_l1 = nn.L1Loss()
criterion_pixel_l = nn.MSELoss()
criterion_pixel_ab = nn.MSELoss()
criterion_style = nn.MSELoss()
criterion_feat = nn.MSELoss()
criterion_texturegan = nn.MSELoss()
return criterion_gan, criterion_pixel_l, criterion_pixel_ab, criterion_style, criterion_feat, criterion_texturegan
def main(args):
#with torch.cuda.device(args.gpu):
layers_map = {'relu4_2': '22', 'relu2_2': '8', 'relu3_2': '13','relu1_2': '4'}
vis = visdom.Visdom(port=args.display_port)
loss_graph = {
"g": [],
"gd": [],
"gf": [],
"gpl": [],
"gpab": [],
"gs": [],
"d": [],
"gdl": [],
"dl": [],
}
# for rgb the change is to feed 3 channels to D instead of just 1. and feed 3 channels to vgg.
# can leave pixel separate between r and gb for now. assume user use the same weights
transforms = get_transforms(args)
if args.color_space == 'rgb':
args.pixel_weight_ab = args.pixel_weight_rgb
args.pixel_weight_l = args.pixel_weight_rgb
rgbify = custom_transforms.toRGB()
train_dataset = ImageFolder('train', args.data_path, transforms)
train_loader = DataLoader(dataset=train_dataset, batch_size=args.batch_size, shuffle=True)
val_dataset = ImageFolder('val', args.data_path, transforms)
indices = torch.randperm(len(val_dataset))
val_display_size = args.batch_size
val_display_sampler = SequentialSampler(indices[:val_display_size])
val_loader = DataLoader(dataset=val_dataset, batch_size=val_display_size, sampler=val_display_sampler)
# renormalize = transforms.Normalize(mean=[+0.5+0.485, +0.5+0.456, +0.5+0.406], std=[0.229, 0.224, 0.225])
feat_model = models.vgg19(pretrained=True)
netG, netD, netD_local = get_models(args)
criterion_gan, criterion_pixel_l, criterion_pixel_ab, criterion_style, criterion_feat,criterion_texturegan = get_criterions(args)
real_label = 1
fake_label = 0
optimizerD = optim.Adam(netD.parameters(), lr=args.learning_rate_D, betas=(0.5, 0.999))
optimizerG = optim.Adam(netG.parameters(), lr=args.learning_rate, betas=(0.5, 0.999))
optimizerD_local = optim.Adam(netD_local.parameters(), lr=args.learning_rate_D_local, betas=(0.5, 0.999))
with torch.cuda.device(args.gpu):
netG.cuda()
netD.cuda()
netD_local.cuda()
feat_model.cuda()
criterion_gan.cuda()
criterion_pixel_l.cuda()
criterion_pixel_ab.cuda()
criterion_feat.cuda()
criterion_texturegan.cuda()
input_stack = torch.FloatTensor().cuda()
target_img = torch.FloatTensor().cuda()
target_texture = torch.FloatTensor().cuda()
segment = torch.FloatTensor().cuda()
label = torch.FloatTensor(args.batch_size).cuda()
label_local = torch.FloatTensor(args.batch_size).cuda()
extract_content = FeatureExtractor(feat_model.features, [layers_map[args.content_layers]])
extract_style = FeatureExtractor(feat_model.features,
[layers_map[x.strip()] for x in args.style_layers.split(',')])
model = {
"netG": netG,
"netD": netD,
"netD_local": netD_local,
"criterion_gan": criterion_gan,
"criterion_pixel_l": criterion_pixel_l,
"criterion_pixel_ab": criterion_pixel_ab,
"criterion_feat": criterion_feat,
"criterion_style": criterion_style,
"criterion_texturegan": criterion_texturegan,
"real_label": real_label,
"fake_label": fake_label,
"optimizerD": optimizerD,
"optimizerD_local": optimizerD_local,
"optimizerG": optimizerG
}
for epoch in range(args.load_epoch, args.num_epoch):
train(model, train_loader, val_loader, input_stack, target_img, target_texture,
segment, label, label_local,extract_content, extract_style, loss_graph, vis, epoch, args)
#break
if __name__ == '__main__':
args = argparser.parse_arguments()
main(args)